Spacecraft system control based on event-triggered forced composite disturbance observer

Abstract

The invention discloses a spacecraft system control method based on an event-triggered forced composite disturbance observer, and the method comprises the steps: building an ideal dynamic model of a spacecraft based on a Clohessy-Wiltshire equation, and carrying out the constraint of the state of the spacecraft; when an event triggering mechanism and an outlier are considered at the same time, an original relative motion model of a spacecraft system is rewritten into a hybrid system form; designing a radial basis function neural network for estimating an unknown continuous nonlinear function in the system; designing a forced composite disturbance observer based on event triggering, wherein the forced composite disturbance observer is used for estimating a system state and multi-source disturbance under an outlier; and designing a composite anti-interference controller based on an event-triggered forced composite interference observer and model prediction control estimation, and constraining the control quantity of the spacecraft. By optimizing the spacecraft system, the control problem of the spacecraft system under the conditions of communication congestion, multi-source uncertainty and input constraint in the prior art is well solved.

Description

Technical Field

[0001] This invention belongs to the field of spacecraft control technology, and specifically relates to a spacecraft system control based on an event-triggered forced composite interference observer. Background Technology

[0002] In recent years, due to the crucial role of spacecraft in critical orbital missions, the precise control of spacecraft systems has been a major focus. Typically, spacecraft transmit their measurement information to information processing terminals (such as space stations or ground stations) via wireless communication equipment and await subsequent control commands. However, during on-orbit servicing, the limited capacity of onboard communication equipment and the inherent multiple-input multiple-output (MIMO) characteristics of spacecraft place a significant burden on the limited communication bandwidth when transmitting large amounts of measurement data. This often leads to problems such as packet loss and transmission delays. Furthermore, signal transmission is highly susceptible to electromagnetic interference and other disturbances, inevitably causing anomalies in the measurement data. Data deviating from the normal trend of the measurement data are called outliers. To address communication bandwidth limitations, existing research mainly employs time-triggered or event-triggered mechanisms to reduce data transmission volume. These methods not only alleviate the communication burden but also lead to phenomena such as system state jitter, making them unsuitable for spacecraft relative motion control missions requiring high precision. In addition, to address the outlier problem, state observers with constraint functions have been proposed.

[0003] Furthermore, considering the complex space environment, spacecraft face multi-source uncertainties during space operations. These uncertainties can be categorized into internal uncertainties (such as elastic disturbances introduced by flexible devices like antennas and solar panels, as well as actuator noise) and external uncertainties from environmental disturbances (including gravitational gradient torque, solar radiation, and geomagnetic torque). From a modeling perspective, these disturbances can generally be classified as modelable disturbances and unmodeled dynamics. To address the adverse effects of these disturbances, disturbance observers have been designed and widely applied in control systems for disturbance estimation. For modelable disturbances, if the disturbance can be described by an exogenous system (modeled disturbance), the disturbance observer can utilize information from the exogenous system to asymptotically stabilize the error dynamics of the observer system. For slow time-varying disturbances that cannot be accurately modeled but are known to have bounded derivatives, researchers have designed extended state observers to observe them. Additionally, for unknown nonlinear terms in the system, radial basis function neural networks are used for estimation. However, existing multi-source uncertainty combined observers estimate the actual output of the system as the observer input, without considering the influence of sensor outliers on the observation results. Ignoring outliers prevents the observer from accurately estimating unknown states and uncertainties in real time using real measurement data. Especially when using high-gain observers, outliers can be amplified, significantly reducing estimation accuracy and even causing closed-loop system interruptions. Therefore, designing a composite disturbance observer capable of observing multi-source uncertainties in the presence of outliers is the motivation for this research. Another factor posing a challenge to precise spacecraft control is input constraints. Input constraints stem from the physical characteristics of the spacecraft actuators themselves. To address input constraint issues, researchers have proposed using model predictive control (MMC) methods to obtain control inputs that satisfy the constraints. However, current MMC methods still haven't explored the impact of communication and multi-source uncertainties. Therefore, designing a MMC method to address the simultaneous impact of communication and multi-source uncertainties on the system is another motivation for this research. Summary of the Invention

[0004] Purpose of the Invention: To overcome the above shortcomings, the purpose of this invention is to provide a spacecraft system control based on an event-triggered forced composite interference observer. By introducing a radial basis function neural network through a static event-triggered mechanism, the unknown nonlinear terms in the system can be estimated, which can effectively solve the communication congestion problem in the prior art. The event-triggered forced composite interference observer can effectively solve the multi-source uncertainty problem. At the same time, a composite anti-interference controller based on model predictive control is also set up to control the spacecraft attitude, thereby effectively solving the control problem of the spacecraft system under input constraints.

[0005] Technical solution: To achieve the above objectives, the present invention provides a spacecraft system control based on an event-triggered forced composite interference observer, comprising the following steps;

[0006] S1): Based on the Clohessy-Wiltshire equations, an ideal dynamic model of the spacecraft is established, and the state of the spacecraft is constrained;

[0007] S2): When considering both event triggering mechanisms and outliers, the original relative motion model of the spacecraft system is rewritten as a hybrid system.

[0008] S3): Design a radial basis function neural network to estimate unknown continuous nonlinear functions in a system;

[0009] Design an event-triggered forced composite disturbance observer to estimate the system state x(t) and the multi-source disturbances d(t) and d under outliers. o (t);

[0010] S4): For spacecraft systems, design a composite anti-interference controller based on event-triggered forced composite interference observer and model predictive control estimation, and constrain the control variables of the spacecraft.

[0011] The spacecraft system control based on the event-triggered forced composite interference observer described in this invention, in step S1), establishes an ideal dynamic model of the spacecraft based on the Clohessy-Wiltshire equations, as detailed below:

[0012]

[0013] in, It refers to the status of the spacecraft system. Is the recipient and Constrained system control inputs; It is an unknown continuous nonlinear function that depends on the system state; It is the output of the spacecraft system; and It is a system matrix with appropriate dimensions; Let x be the derivative.

[0014] The spacecraft system control based on the event-triggered forced composite interference observer described in this invention, when considering both the event triggering mechanism and outliers, rewrites the original relative motion model of the spacecraft system corresponding to formulas (1)-(2) in S2) into a hybrid system form, as shown below:

[0015]

[0016] y(tk )=x1(t k )+β(t k )v(t k (4)

[0017] Among them, t k y(t) represents the event trigger time. k ) represents the wireless telemetry data transmitted at the moment the event is triggered; v(t) k ) represents an outlier in wireless transmission; β(t) k ) indicates whether an outlier exists; when β(t) k When β(t) = 1, it indicates the existence of an outlier. k When ) = 0, it indicates that there are no outliers;

[0018] in, It is to satisfy External disturbances, where d d >0 is a known constant;

[0019] Generated by the following exogenous systems:

[0020]

[0021] in, It is a perturbation state d o (t); and It is a known matrix.

[0022] In the spacecraft system control based on an event-triggered forced composite interference observer described in this invention, a radial basis function neural network is designed in step S3) to estimate the unknown continuous nonlinear function f(x(t)) in the system, as follows:

[0023] Defined in compact sets Continuous nonlinear function on It can be approximated by the following radial basis function neural network:

[0024] f(x(t))=θ *T η(x(t))+π(t)

[0025] in, and These are the ideal weight vector, the vector-valued function, and the network reconstruction error, respectively; furthermore, η(x(t)) is norm-bounded, satisfying... Where ξ>0 and It is a known constant;

[0026] π(t) is also continuously differentiable, where ||π(t)|| ≤ π m and π m and π d It is a known constant;

[0027] Formulas (3)-(4) for the hybrid system can be rewritten as:

[0028]

[0029] y(t k )=x1(t k )+β(t k )v(t k (6)

[0030] in, and θ(t) is θ * The estimate;

[0031] Then, formulas (5)-(6) can be rewritten as:

[0032]

[0033] y(t k )=x1(t k )+β(t k )v(t k (8)

[0034] in,

[0035]

[0036] in, and These are the relative position and velocity components on the three coordinate axes: x-axis, y-axis, and z-axis; x T Indicates the transpose of x;

[0037] Let A be a matrix with dimension m×n; A > 0 indicates that A is a positive definite matrix.

[0038] A < 0 indicates that A is a negative definite matrix; I indicates that the identity matrix has appropriate dimensions.

[0039] The spacecraft system control based on the event-triggered forced composite disturbance observer described in this invention, in S3), designs an event-triggered forced composite disturbance observer to estimate the system state x(t) and the multi-source disturbances d(t) and d under outliers. o (t):

[0040]

[0041]

[0042]

[0043] in, It is an estimate of X(t).

[0044] L is the gain matrix of an event-triggered forced composite interference observer, satisfying L = [L1 L2 L3] T Where L1, L2, and L3 are given parameters; ρ(t) is the estimate based on the between-sample term. The variables; α1>0, α2>0, γ3>0 are adjustable parameters of the event-triggered forced composite disturbance observer;

[0045] Where E(t) is the event triggering mechanism condition, which is as follows:

[0046]

[0047] in,

[0048] C = [I000],

[0049] α e >0 and ∈0>0 are two adjustable positive parameters;

[0050] E(t) = 1 and E(t) = 0 represent that the event was triggered and the event was not triggered, respectively;

[0051] sat χ (·) is a saturation function, which has a non-negative dynamic variable χ(t) as its upper limit;

[0052] sat χ The specific expression for (ρ(t)-z1(t)) is:

[0053] sat x (ρ(t)-z1(t))=[sat χ (ρ1(t)-z 11 (t)), sat χ (ρ2(t)-z 12 (t)), sat χ (ρ3(t)-z 13 (t))] T

[0054] Among them, sat χ (ρ i (t)-z 1i (t))=max{-χ(t),min{χ(t),ρ i (t)-z 1i (t)}},

[0055] i = 1, 2, 3; Definition This allows the dynamics of the estimation error to be expressed by formulas (4)-(9).

[0056] The result is:

[0057]

[0058] In the spacecraft system control based on an event-triggered forced composite interference observer described in this invention, in step S4), an event-triggered forced composite interference observer and a module are designed for the spacecraft system.

[0059] Composite anti-interference controller for predictive control estimation:

[0060] The nominal system representation of system model formula (7) is:

[0061]

[0062] in, It is the nominal state of the system according to formula (5); It is formula (5) system

[0063] Standardized control input, among which It is the optimal control signal generated by the predictive control scheme;

[0064] Then, the composite anti-interference controller of the spacecraft system is designed as follows:

[0065]

[0066] in, It is the optimal control signal generated by the model predictive control scheme; It is an estimate of x(t); K is the value that makes x(t) equal to x(t). Stable feedback gain; It is d o Estimate (t);

[0067] In order to obtain This introduces a new constrained optimization problem for spacecraft systems.

[0068]

[0069] Constraints

[0070]

[0071] Where ||·|| denotes the Euclidean norm; 0 denotes a matrix of appropriate dimension filled with zeros; 1 denotes a matrix of appropriate dimension filled with one; x(s|t k ) indicates from time tk Prediction from time s to x; Indicates from time t k Prediction of the state up to time s; Indicates from time t k The model predicts the control input signal up to time s; δ is the sampling period, δ = t k+1 -t k T is the prediction range that satisfies T≥2δ;

[0072] definition

[0073]

[0074] Where σ > 0 is a given constant; Q > 0 and R > 0 are symmetric matrices representing weight matrices with appropriate dimensions; P > 0 is a symmetric matrix representing the terminal penalty matrix to be designed; in Obtained from the following formula:

[0075]

[0076] As can be seen from the above technical solution, the present invention has the following beneficial effects:

[0077] 1. The spacecraft system control based on an event-triggered forced composite interference observer described in this invention, through a static event triggering mechanism, introduces a radial basis function neural network to estimate the unknown nonlinear terms in the system, which can effectively solve the communication congestion problem in the prior art; through further optimization of the spacecraft system, it solves the control problem of the spacecraft system under communication congestion, multi-source uncertainty and input constraints in the prior art.

[0078] 2. In addition, this invention also designs an event-triggered forced composite disturbance observer algorithm to simultaneously obtain the estimation of system state and multi-source disturbances under outliers, thereby significantly improving the accuracy of the estimation.

[0079] 3. This invention also proposes a composite anti-interference controller based on pipeline-MPC to suppress multi-source uncertainties, solve the outlier problem of the system being affected by communication and multi-source uncertainties at the same time, and thus better control the spacecraft. Attached Figure Description

[0080] Figure 1 A flowchart illustrating the spacecraft system control based on an event-triggered forced composite interference observer in this invention;

[0081] Figure 2 A state curve diagram of a spacecraft in an embodiment provided by the present invention. Detailed Implementation

[0082] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0083] Example 1

[0084] like Figures 1 to 2 The spacecraft system control based on an event-triggered forced composite interference observer, as shown, includes the following steps:

[0085] S1): Based on the Clohessy-Wiltshire equations, an ideal dynamic model of the spacecraft is established, and the state of the spacecraft is constrained;

[0086] S2): When considering both event triggering mechanisms and outliers, the original relative motion model of the spacecraft system is rewritten as a hybrid system.

[0087] S3): Design a radial basis function neural network to estimate unknown continuous nonlinear functions in a system;

[0088] Design an event-triggered forced composite disturbance observer to estimate the system state x(t) and the multi-source disturbances d(t) and d under outliers. o (t);

[0089] S4): For spacecraft systems, design a composite anti-interference controller based on event-triggered forced composite interference observer and model predictive control estimation, and constrain the control variables of the spacecraft.

[0090] In this embodiment, the spacecraft system control based on the event-triggered forced composite interference observer in S1) establishes an ideal dynamic model of the spacecraft based on the Cloheessy-Wiltshire equations, as detailed below:

[0091]

[0092] in, It refers to the status of the spacecraft system. Is the recipient and Constrained system control inputs; It is an unknown continuous nonlinear function that depends on the system state; It is the output of the spacecraft system; and It is a system matrix with appropriate dimensions; Let x be the derivative.

[0093] Due to the long distance between the ground station, the tracking spacecraft, and the target spacecraft node, and the large amount of telemetry data, if all measurement data is transmitted through the satellite-ground link, data congestion is likely to occur, leading to delays or data loss. To avoid data congestion, an event-triggered mechanism will be adopted, meaning that telemetry data will only be transmitted when the transmission data meets the set triggering conditions. Compared with the traditional time-triggered method, this will greatly reduce the amount of data transmitted. In addition, the transmission process of telemetry data is highly susceptible to interference from cosmic rays and atmospheric electromagnetic signals, causing fluctuations in the original measurement data, known as outliers. When considering both the event-triggered mechanism and outliers, the original relative motion model of the spacecraft system corresponding to formulas (1)-(2) in S2) is rewritten as a hybrid system, as shown below:

[0094]

[0095] y(t k )=x1(t k )+β(t k )v(t k (4)

[0096] Among them, t k y(t) represents the event trigger time. k ) represents the wireless telemetry data transmitted at the moment the event is triggered; v(t) k ) represents an outlier in wireless transmission; β(t) k ) indicates whether an outlier exists; when β(t) k When β(t) = 1, it indicates the existence of an outlier. k When ) = 0, it indicates that there are no outliers;

[0097] In addition, β(t) k Prob{β(t)} satisfies a Bernoulli distribution. k )=0}=1-β and Prob{β(t) k )=1}=β, where the constant β∈[0,1), and Prob represents the probability.

[0098] in, It is to satisfy External disturbances, where d d >0 is a known constant;

[0099] Generated by the following exogenous systems:

[0100]

[0101] in, It is a perturbation state d o (t); and It is a known matrix;

[0102] A composite anti-interference controller based on model predictive control is used to simultaneously obtain estimates of system state and multi-source disturbances under outlier conditions, so as to suppress multi-source uncertainty and solve the outlier problem of system being affected by communication and multi-source uncertainty at the same time.

[0103] In the spacecraft system control based on the event-triggered forced composite interference observer described in this embodiment, a radial basis function neural network is designed in step S3) to estimate the unknown continuous nonlinear function f(x(t)) in the system, as detailed below:

[0104] Defined in compact sets Continuous nonlinear function on It can be approximated by the following radial basis function neural network:

[0105] f(x(t))=θ *T η(x(t))+π(t)

[0106] in, and These are the ideal weight vector, the vector-valued function, and the network reconstruction error, respectively; furthermore, η(x(t)) is norm-bounded, satisfying... Where ξ>0 and It is a known constant;

[0107] π(t) is also continuously differentiable, where ||π(t)||≤π m and π m and π d It is a known constant;

[0108] The hybrid system of formulas (3)-(4) can be rewritten as:

[0109]

[0110] y(t k )=x1(t k )+β(t k )v(t k (6)

[0111] in, and θ(t) is θ * The estimate;

[0112] Then, the system corresponding to formulas (5)-(6) can be rewritten as:

[0113]

[0114] y(tk )=x1(t k )+β(t k )v(t k (8)

[0115] in,

[0116]

[0117] in, and These are the relative position and velocity components on the three coordinate axes: x-axis, y-axis, and z-axis; x T Indicates the transpose of x;

[0118] Let A be a matrix with dimension m×n; A>0 indicates that A is a positive definite matrix.

[0119] A<0 indicates that A is a negative definite matrix; I represents an identity matrix of appropriate dimension. During implementation, a static event-triggered mechanism is used to introduce a radial basis function neural network to estimate unknown nonlinear terms in the system, effectively addressing the communication congestion problem in existing technologies.

[0120] In this embodiment, the spacecraft system control based on the event-triggered forced composite disturbance observer (S3) is designed to estimate the system state x(t) and the multi-source disturbances d(t) and d under outliers. o (t):

[0121]

[0122] in, It is an estimate of X(t).

[0123] L is the gain matrix of an event-triggered forced composite interference observer, satisfying L = [L1 L2 L3] T Where L1, L2, and L3 are given parameters; ρ(t) is the estimate based on the between-sample term. The variables; α1 > 0, α2 > 0, γ3 > 0 are event-triggered forced complex variables.

[0124] Adjustable parameters of the combined interference observer;

[0125] Where E(t) is the event triggering mechanism condition, which is as follows:

[0126]

[0127] in,

[0128] C = [I000],

[0129] α e >0 and ∈0>0 are two adjustable positive parameters;

[0130] E(t) = 1 and E(t) = 0 represent that the event was triggered and the event was not triggered, respectively;

[0131] sat χ (·) is a saturation function, which has a non-negative dynamic variable χ(t) as its upper limit;

[0132] sat χ The specific expression for (ρ(t)-z1(t)) is:

[0133] sat χ (ρ(t)-z1(t))=[sat χ (ρ1(t)-z 11 (t)), sat χ (ρ2(t)-z 12 (t)), sat χ (ρ3(t)-z 13 (t))] T

[0134] Among them, sat χ (ρ i (t)-z 1i (t))=max{-χ(t),min{χ((t),ρ i (t)-z 1i (t)}},

[0135] i = 1, 2, 3; Definition This allows the dynamics of the estimation error to be expressed by formulas (4)-(9).

[0136] The result is:

[0137]

[0138] In the spacecraft system control based on the event-triggered forced composite interference observer described in this embodiment, in step S4), a composite anti-interference controller based on the event-triggered forced composite interference observer and the model predictive control estimation is designed for the spacecraft system.

[0139] The nominal system representation of the system model of formula (7) is:

[0140]

[0141] in, It is the nominal state of the system according to formula (5); It is the nominal control input of the system in formula (5), where It is the optimal control signal generated by the predictive control scheme;

[0142] Then, the composite anti-interference controller of the spacecraft system is designed as follows:

[0143]

[0144] in, It is the optimal control signal generated by the model predictive control scheme; It is an estimate of x(t); K is the value that makes x(t) equal to x(t). Stable feedback gain; It is d o Estimate (t);

[0145] In order to obtain This introduces a new constrained optimization problem for spacecraft systems.

[0146]

[0147] Constraints

[0148]

[0149] Where ||·|| denotes the Euclidean norm; 0 denotes a matrix of appropriate dimension filled with zeros; 1 denotes a matrix of appropriate dimension filled with one; x(s|t k ) indicates from time t k Prediction from time s to x; Indicates from time t k Prediction of the state up to time s; Indicates from time t k The model predicts the control input signal up to time s; δ is the sampling period, δ = t k+1 -t k T is the prediction range that satisfies T≥2δ;

[0150] definition

[0151]

[0152] Where σ > 0 is a given constant; Q > 0 and R > 0 are symmetric matrices representing weight matrices with appropriate dimensions; P > 0 is a symmetric matrix representing the terminal penalty matrix to be designed; in Obtained from the following formula:

[0153]

[0154] The present invention will now be described in conjunction with specific embodiments:

[0155] This embodiment employs the spacecraft system control method based on the event-triggered forced composite interference observer described above. It should be noted that in this embodiment, in formulas (5)-(11),

[0156] β = 0.6

[0157] in, The constraints on the control input are Where i = 1, 2, 3. d o The parameters of (t) are as follows:

[0158]

[0159] Among them, the perturbation d(t) and the nonlinear term f(x(t)) are considered to be

[0160]

[0161] Furthermore, the parameters for the event-triggered forced composite interference observer in equations (9)-(11) are given:

[0162] α1=15, α2=0.01, α e =0.4, ∈0=10, L3 = 10I 3×3 In this embodiment, see Figure 2 The state curves of the spacecraft are given, where x i (t)(i=1,2,3,4,5,6) represents the state of the spacecraft. It can be seen that under the control method designed in this invention, the state converges smoothly.

[0163] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements can be made without departing from the principle of the present invention, and these improvements should also be considered within the scope of protection of the present invention.

Claims

1. A spacecraft system control based on an event-triggered forced composite interference observer, characterized in that: include: The following steps; S1): Based on the Clohessy-Wiltshire equations, an ideal dynamic model of the spacecraft is established, and the state of the spacecraft is constrained; S2): When considering both event triggering mechanisms and outliers, the original relative motion model of the spacecraft system is rewritten as a hybrid system. S3): Design a radial basis function neural network to estimate unknown continuous nonlinear functions in a system; Design an event-triggered forced composite disturbance observer to estimate the system state x(t) and the multi-source disturbances d(t) and d under outliers. o (t); S4): For spacecraft systems, design a composite anti-interference controller based on event-triggered forced composite interference observer and model predictive control estimation, and constrain the control variables of the spacecraft.

2. The spacecraft system control based on the event-triggered forced composite interference observer according to claim 1, characterized in that: In S1), an ideal dynamic model of the spacecraft is established based on the Clohessy-Wiltshire equations, as follows: in, It refers to the status of the spacecraft system. Is the recipient and Constrained system control inputs; It is an unknown continuous nonlinear function that depends on the system state; It is the output of the spacecraft system; and It is a system matrix with appropriate dimensions; Let x be the derivative.

3. The spacecraft system control based on the event-triggered forced composite interference observer according to claim 2, characterized in that: When both event triggering mechanisms and wild values ​​are considered, the original relative motion model of the spacecraft system corresponding to formulas (1)-(2) in S2) is rewritten as a hybrid system, as follows: y(t k )=x1(t k )+β(t k )v(t k ) (4) Among them, t k y(t) represents the event trigger time. k ) represents the wireless telemetry data transmitted at the moment the event is triggered; v(t) k ) represents an outlier in wireless transmission; β(t) k ) indicates whether an outlier exists; when β(t) k When β(t) = 1, it indicates the existence of an outlier. k When ) = 0, it indicates that there are no outliers; It is to satisfy External disturbances, where d d >0 is a known constant; Generated by the following exogenous systems: d o (t)=Vw(t); in, It is a perturbation state d o (t); and It is a known matrix.

4. The spacecraft system control based on the event-triggered forced composite interference observer according to claim 3, characterized in that: In S3), a radial basis function neural network is designed to estimate the unknown continuous nonlinear function f(x(t)) in the system, as follows: Defined in compact sets Continuous nonlinear function on It can be approximated by the following radial basis function neural network: f(x(t))=θ *T η(x(t))+π(t) in, and These are the ideal weight vector, the vector-valued function, and the network reconstruction error, respectively; furthermore, η(x(t)) is norm-bounded, satisfying... Where ξ>0 and It is a known constant; π(t) is also continuously differentiable, where ||π(t)||≤π m and π m and π d It is a known constant; Formulas (3)-(4) for the hybrid system can be rewritten as: y(t k )=x1(t k )+β(t k )v(t k ) (6) in, and θ(t) is θ * The estimate; Then, formulas (5)-(6) can be rewritten as: y(t k )=x1(t k )+β(t k )v(t k ) (8) in, in, and These are the relative position and velocity components on the three coordinate axes: x-axis, y-axis, and z-axis; x T Indicates the transpose of x; Let A be a matrix with dimension m×n; A > 0 indicates that A is a positive definite matrix. A < 0 indicates that A is a negative definite matrix; I indicates that the identity matrix has appropriate dimensions.

5. The spacecraft system control based on the event-triggered forced composite interference observer according to claim 4, characterized in that: In S3), an event-triggered forced composite disturbance observer is designed to estimate the system state x(t) and the multi-source disturbances d(t) and d under outliers. o (t): in, It is an estimate of X(t). L is the gain matrix of an event-triggered forced composite interference observer, satisfying L = [L1 L2 L3] T Where L1, L2, and L3 are given parameters; ρ(t) is the estimate based on the between-sample term. The variables; α1>0, α2>0, γ3>0 are adjustable parameters of the event-triggered forced composite disturbance observer; Where E(t) is the event triggering mechanism condition, which is as follows: in, α e >0 and ∈0>0 are two adjustable positive parameters; E(t) = 1 and E(t) = 0 represent that the event was triggered and the event was not triggered, respectively; sat x (·) is a saturation function, which has a non-negative dynamic variable x(t) as its upper limit; sat x The specific expression for (ρ(t)-z1(t)) is: sat x (ρ(t)-z1(t))=[sat x (ρ1(t)-z 11 (t)),sat x (ρ2(t)—z 12 (t)),sat x (ρ3(t)-z 13 (t))] T where, sat x (ρ i (t) - z 1i (t)) = max{-x(t), min{x(t), ρ i (t) - z 1i (t)}, i = 1, 2, 3; definition The dynamics of the estimation error can be obtained through formulas (4)-(9):

6. The spacecraft system control based on the event-triggered forced composite interference observer according to claim 5, characterized in that: In S4), for the spacecraft system, a composite anti-interference controller based on event-triggered forced composite interference observer and model predictive control estimation is designed: The nominal system representation of the system model corresponding to formula (7) is: in, It is the nominal state of the system according to formula (5); It is the nominal control input of the system in formula (5), where It is the optimal control signal generated by the predictive control scheme; Then, the composite anti-interference controller of the spacecraft system is designed as follows: in, It is the optimal control signal generated by the model predictive control scheme; It is an estimate of x(t); K is the value that makes x(t) equal to x(t). Stable feedback gain; It is d o Estimate (t); In order to obtain This introduces a new constrained optimization problem for spacecraft systems. Constraints Where ||·|| denotes the Euclidean norm; 0 denotes a matrix of appropriate dimension filled with zeros; 1 denotes a matrix of appropriate dimension filled with one; x(s|t k ) indicates from time t k Prediction from time s to x; Indicates from time t k Prediction of the state up to time s; Indicates from time t k The model predicts the control input signal up to time s; δ is the sampling period, δ = t k+1 -t k T is the prediction range that satisfies T≥2δ; definition Where σ > 0 is a given constant; Q > 0 and R > 0 are symmetric matrices representing weight matrices with appropriate dimensions; P > 0 is a symmetric matrix representing the terminal penalty matrix to be designed. in Obtained from the following formula:

Images